Research Scientist – Neural 3D Objects

The role

You will own the architecture for generating 3D objects from video. You will build models to recover geometry from egocentric footage to create simulatable assets with physical properties.

What you will do

• Neural 3D reconstruction from monocular or egocentric video
• Generative refinement to complete and clean up partial reconstructions
• Developing assets that carry physical properties and rigging for simulation
• Contact-aware reconstruction using hand contact as a signal
• Working on deformable objects that change shape
• Building research prototypes and handing them to the integrations researcher
• Publishing research where it makes sense

Required qualifications

• Strong 3D / neural-reconstruction research in neural implicit or explicit 3D, 4D reconstruction, or generative 3D
• Deep-learning depth to design and train models and reproduce frontier papers
• 3D geometry foundation in structure-from-motion, multi-view geometry, meshes, and differentiable rendering
• Research taste to find tractable decomposition of hard, open problems
• Engineering ability to prototype fast in clean Python / PyTorch

Nice to have

• Generative 3D (diffusion / feed-forward 3D-generation families)
• Differentiable rendering, NeRF, or Gaussian splatting
• 4D / dynamic-scene reconstruction from video
• Physics simulation and rigging
• Hand-object interaction
• Publications at top vision or graphics venues

About the company

Mecka AI is building the data infrastructure layer for robotics and embodied AI. They design and operate global systems for data capture, data labeling, and hardware-enabled workflows for AI labs and robotics companies.